The conundrum of forest expansion after the last ice age

How fast the Northern Hemisphere forest macro ecosystem tracks strongly warming climates such as projected for the near future is largely unknown. In a recent study published in Nature Communication, scientists investigated the Northern Hemisphere forest expansion after the last glacial maximum. They compared a new synthesis of pollen-based reconstructions and a climate model simulation of the last 22,000 years. In the process, they discovered a difference of several thousand years in the expansion of forests. This conundrum challenges the paleo-climate community. Shortcomings in the model and the reconstructions could both contribute to this mismatch, but can technically not been disentangled so far, leaving the underlying causes unresolved.

Credit: Stefan Kruse/AWI

The authors of the study — Anne Dallmeyer, Thomas Kleinen, Martin Claussen (Max Planck Institute for Meteorology und CEN, Universität Hamburg) and researchers from the Alfred-Wegener Institute, University Potsdam, University Heidelberg, University Tübingen, and the Chinese Academy of Sciences — have shown that the climate simulated by the Earth system model MPI-ESM1.2 is consistent with pollen-independent climate reconstructions for the last 22,000 years. Thus, possible biases in the simulated climate are not the main driver of the model-data discrepancy. The assignment of individual plant genera to different macro-ecosystems is not straightforward. It relies to some extent on subjective and individual choices, which may lead to uncertainties in the reconstructions. The main reason, however, may be the fact that the simulated vegetation follows the simulated climate within decades and is thus in quasi-equilibrium with the climate on the analyzed time-scale, while the real forest has to migrate from one region to others. This migration involves in turn slow ecological processes such as the development of soils on bare rocks that have been exposed during the deglaciation. These processes that slow down the migration of forest are so far not implemented in Earth system models. Both the direct response of the vegetation to a changing climate in the model and the slow migration processes of individual tree species imply that the discovered model-data mismatch points at a multi-millennial disequilibrium of the Northern Hemisphere forest dynamic and a changing climate. The mismatch is particularly pronounced in Asia.

Change in Northern Hemisphere forest coverage. Simulated (black) and reconstructed (red) mean forest cover on the Northern Hemisphere, north of 30oN, for the last 22,000 years and the respective uncertainty in forest cover (grey and red shadings). The late-glacial period (22-18 ka) and the Holocene are shaded.

The consequences of the model-data discrepancy are twofold. Pollen records are still the most wide-spread terrestrial archive for climate and vegetation reconstructions and are commonly used to evaluate the performance of Earth system models for past climate states, also based on time-slice experiments. If the disequilibrium is real, the usage of pollen records for model validation should be critically reassessed. Second, the results by Dallmeyer et al. imply that Northern Hemisphere forests may be responding much slower to ongoing climate changes than Earth system models predict. This is a major source of uncertainty in the prediction of how climate change will affect changes in forests and biodiversity.

The study was conducted as part of the BMBF-funded PalMod project ( and the CLICCS cluster of excellence (


Original paper:
Dallmeyer, A., Kleinen, T., Claussen, M., Weitzel, N., Cao, X. & Herzschuh, U. (2022). The deglacial forest conundrum. Nature Communications, 13: 6035. doi:10.1038/s41467-022-33646-6


Dr. Anne Dallmeyer
Max Planck Institute for Meteorology
anne.dallmeyer@we dont want

Prof. Dr. Martin Claussen
Max Planck Institute for Meteorology
martin.claussen@we dont want

Dr. Thomas Kleinen
Max Planck Institute for Meteorology
thomas.kleinen@we dont want